Oxygen homeostasis is an essential cellular and systemic function; hypoxia leads to metabolic demise, but this must be balanced by the risk of oxidative damage to cellular lipids, nucleic acids, and proteins resulting from hyperoxia. As a result, cellular and systemic oxygen concentrations are tightly regulated via response pathways that affect the activity and expression of a multitude of cellular proteins. This balance is disrupted in heart disease, cancer, cerebrovascular disease, and chronic obstructive pulmonary disease (Semenza, Genes Dev., 2000, 14, 1983-1991)(Semenza, G., 2001, Trends Mol. Med., 7, 345-350. Cells are typically cultured in the laboratory at an ambient oxygen concentration of 21%, but cells in the human body are exposed to much lower oxygen concentrations ranging from 16% in the lungs to less than 6% in most other organs of the body—often significantly less in tumors. Semenza G., 2001, Trends Mol. Med., 7, 345-350.
Solid tumor growth depends on a continuous supply of oxygen and nutrients through neovascularization (angiogenesis). Tumors often become hypoxic, often because new blood vessels are aberrant and have poor blood flow. Cancer cells make adaptive changes that allow them to proliferate even at hypoxia. These changes include an increase in glycolysis and an increase in production of angiogenic factors. Hypoxia in tumors is associated with resistance to radio-and chemotherapy, and thus is an indicator of poor survival.
The transcriptional complex, hypoxia inducible factor (HIF), is a key regulator of oxygen homeostasis. Hypoxia induces the expression of genes participating in many cellular and physiological processes, including oxygen transport and iron metabolism, erythropoiesis, angiogenesis, glycolysis and glucose uptake, transcription, metabolism, pH regulation, growth-factor signaling, response to stress and cell adhesion. These gene products participate in either increasing oxygen delivery to hypoxic tissues or activating an alternative metabolic pathway (glycolysis) which does not require oxygen. Hypoxia-induced pathways, in addition to being required for normal cellular processes, can also aid tumor growth by allowing or aiding angiogenesis, immortalization, genetic instability, tissue invasion and metastasis (Harris, Nat. Rev. Cancer, 2002, 2, 38-47; Maxwell et al., Curr. Opin. Genet. Dev., 2001, 11, 293-299).
HIF is a heterodimer composed of an alpha subunit complexed with a beta subunit, both of which are basic helix-loop-helix transcription factors. The beta subunit of HIF is a constitutive nuclear protein. The alpha subunit is the regulatory subunit specific to the oxygen response pathway, and can be one of three subunits, HIF1α, 2 alpha or 3 alpha (HIF-1α, HIF-2α and HIF-3α, respectively) (Maxwell et al., Curr. Opin. Genet. Dev., 2001, 11, 293-299; Safran and Kaelin, J. Clin. Invest., 2003, 111, 779-783).
The transcription factor hypoxia-inducible factor 1 (HIF-1) plays an essential role in homeostatic responses to hypoxia by binding to the DNA sequence 5′-TACGTGCT-3′ and activating the transcription of dozens of genes in vivo under hypoxic conditions (Wang and Semenza, J. Biol. Chem., 1995, 270, 1230-1237). These gene products participate in either increasing oxygen delivery to hypoxic tissues or activating an alternative metabolic pathway (glycolysis) which does not require oxygen. This list includes: aldolase C, enolase 1, glucose transporter 1, glucose transporter 3, glyceraldehyde-3-phosphate dehydrogenase, hexokinase 1, hexokinase 2, insulin-like growth factor-2 (IGF-2), IGF binding protein 1, IGF binding protein 3, lactate dehydrogenase A, phosphoglycerate kinase 1, pyruvate kinase M, p21, transforming growth factor B3, ceruloplasmin, erythropoietin, transferrin, transferrin receptor, alb-adrenergic receptor, adrenomedullin, endothelin-1, heme oxygenase 1, nitric oxide synthase 2, plasminogen activator inhibitor 1, vascular endothelial growth factor (VEGF), VEGF receptor FTL-1, and p35 (Semenza, Genes Dev., 2000, 14, 1983-1991). Expression of HIF1α is also sensitive to oxygen concentration: increased levels of protein are detected in cells exposed to 1% oxygen and these decay rapidly upon return of the cells to 20% oxygen (Wang et al., Proc. Natl. Acad. Sci. U.S.A., 1995, 92, 5510-5514).
Hypoxia-inducible factor-1 alpha is a heterodimer composed of a 120 kDa alpha subunit complexed with a 91 to 94 kDa beta subunit, both of which contain a basic helix-loop-helix (Wang and Semenza, J. Biol. Chem., 1995, 270, 1230-1237). The gene encoding hypoxia-inducible factor-1 alpha (HIF1α, also called HIF-1 alpha, HIF1A, HIF-1A, HIF1-A, and MOP1) was cloned in 1995 (Wang et al., Proc. Natl. Acad. Sci. U.S.A., 1995, 92, 5510-5514). A nucleic acid sequence encoding HIF1α is disclosed and claimed in U.S. Pat. No. 5,882,914, as are expression vectors expressing the recombinant DNA, and host cells containing said vectors (Semenza, 1999).
HIF1α expression and HIF-1 transcriptional activity are precisely regulated by cellular oxygen concentration. The beta subunit is a constitutive nuclear protein, while the alpha subunit is the regulatory subunit. HIF1α mRNA is expressed at low levels in tissue culture cells, but it is markedly induced by hypoxia or ischemia in vivo (Yu et al., J. Clin. Invest., 1999, 103, 691-696). HIF1α protein is negatively regulated in non-hypoxic cells by ubiquitination and proteasomal degradation (Huang et al., Proc. Natl. Acad. Sci. U.S.A., 1998, 95, 7987-7992). Under hypoxic conditions, the degradation pathway is inhibited, HIF1α protein levels increase dramatically, and the fraction that is ubiquitinated decreases. HIP1α then translocates to the nucleus and dimerizes with a beta subunit (Sutter et al., Proc. Natl. Acad. Sci. U.S.A., 2000, 97, 4748-4753).
A natural antisense transcript that is complementary to the 3′ untranslated region of HIF1α mRNA has been discovered and is named “aHIF” (Thrash-Bingham and Tartof, J. Natl. Cancer Inst., 1999, 91, 143-151). This is the first case of overexpression of a natural antisense transcript exclusively associated with a specific human malignant disease. aHIF is specifically overexpressed in nonpapillary clear-cell renal carcinoma under both normoxic and hypoxic conditions, but not in papillary renal carcinoma. Although aHIF is not further induced by hypoxia in nonpapillary disease, it can be induced in lymphocytes where there is a concomitant decrease in HIF1α mRNA.
HIF1α plays an important role in promoting tumor progression and is overexpressed in common human cancers, including breast, colon, lung, and prostate carcinoma. Overexpression of HIFs is sometimes observed in cancers, such as clear cel renal cell carcinoma, even at normoxia. Mutations that inactivate tumor suppressor genes or activate oncogenes have, as one of their consequences, upregulation of HIF1α activity, either through an increase in HIF1α protein expression, HIF1α transcriptional activity, or both (Semenza, Pediatr. Res., 2001, 49, 614-617).
Until a tumor establishes a blood supply, the hypoxic conditions limit tumor growth. Subsequent increases in HIF1α activity result in increased expression of target genes such as vascular endothelial growth factor (VEGF). VEGF expression is essential for vascularization and the establishment of angiogenesis in most solid tumors (Iyer et al., Genes Dev., 1998, 12, 149-162). A significant association between hypoxia-inducible factor-1 alpha, VEGF overexpression and tumor grade is also seen in human glioblastoma multiforme, the highest grade glioma in which mean patient survival time is less than one year. The rapidly proliferating tumor outgrows its blood supply, resulting in extensive necrosis, and these regions express high levels of HIF1α protein and VEGF mRNA, suggesting a response of the tumor to hypoxia (Zagzag et al., Cancer, 2000, 88, 2606-2618).
The action of the von Hippel-Landau (VHL) tumor suppressor gene product is implicated in hypoxic gene regulation, in both normal and diseased cells. Individuals with VHL disease are predisposed to renal cysts, clear cell renal carcinoma, phaeochromocytoma, haemangioblastomas of the central nervous system, angiomas of the retina, islet cell tumors of the pancreas, and endolymphatic sac tumors (Pugh and Ratcliffe, Semin. Cancer. Biol., 2003, 13, 83-89). The VHL gene product participates in ubiquitin-mediated proteolysis by acting as the recognition component of the E3-ubiquitin ligase complex involved in the degradation of hypoxia-inducible factor alpha subunits (Cockman et al., J. Biol. Chem., 2000, 275, 25733-25741; Ohh et al., Nat. Cell Biol., 2000, 2, 423-427). In normal cells, VHL/HIF complexes form and target HIF alpha subunits for destruction (Maxwell et al., Nature, 1999, 399, 271-275). This is proposed to occur through hydroxylation of the oxygen-dependent domain of HIF2α and subsequent recognition by the VHL gene product, as recognition of a homologous oxygen-dependent domain is the mechanism by which the VHL protein recognizes HIF1α (Maxwell et al., Nature, 1999, 399, 271-275). HIF2α is in fact hydroxylated by the enzyme prolyl 4-hydroxylases in vitro (Hirsila et al., J. Biol. Chem., 2003).
The p53 tumor suppressor also targets HIF1α for degradation by the proteasome. Loss of p53 activity occurs in the majority of human cancers and indicates that amplification of normal HIF1α levels contributes to the angiogenic switch during tumorigenesis (Ravi et al., Genes Dev., 2000, 14, 34-44).
A mouse model of pulmonary hypertension has shown that local inhibition of HIF1α activity in the lung might represent a therapeutic strategy for treating or preventing pulmonary hypertension in at risk individuals. In pulmonary hypertension hypoxia-induced vascular remodeling leads to decreased blood flow, which leads to progressive right heart failure and death. This hypoxia-induced vascular remodeling is markedly impaired in mice that are partially HIF1α deficient (Yu et al., J. Clin. Invest., 1999, 103, 691-696). Decreased vascular density and retarded solid tumor growth is also seen in mouse embryonic stem cells which are deficient for HIF1α (Ryan et al., Embo J, 1998, 17, 3005-3015).
During hypoxia, cells shift to a glycolytic metabolic mode for their energetic needs and HIF1α is known to upregulate the expression of many glycolytic genes. HIF1α may play a pivotal role in the Warburg effect in tumors, a paradoxical situation in which tumor cells growing under normoxic conditions show elevated glycolytic rates, which enhances tumor growth and expansion. HIF1α mediates the expression of 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase-3, a gene whose protein product maintains levels of the key regulator of glycolytic flux, fructose-2,6-bisphosphate (Minchenko et al., J. Biol. Chem., 2001, 14, 14).
Currently, there are no known therapeutic agents which effectively inhibit the synthesis of HIF1α and to date, investigative strategies aimed at modulating HIF1α function have involved the use of antisense expression vectors and oligonucleotides. These studies have served to define the involvement of HIF1α in disease progression and to identify novel roles of HIF1α in vivo including unique roles for HIF1α as a transcription factor under non-hypoxic conditions and as an inhibitor of gene expression.
Gene transfer of an antisense HIF1α plasmid has been shown to enhance the efficacy of cancer immunotherapy. Antisense therapy was shown to slow, but not eradicate, the growth of EL-4 tumors established in mice. In addition, endogenously expression of HIF1α was almost completely inhibited in these tumors. When antisense therapy was combined with T-cell costimulator B7-1 immunotherapy, the tumors completely and rapidly regressed within 1 week. Furthermore, when these tumor-free mice were rechallenged with EL-4 cells, no tumors emerged, indicating that systemic antitumor immunity had been achieved (Sun et al., Gene Ther., 2001, 8, 638-645).
Activation of HIF1α is thought to aggravate heart failure by upregulation of cardiac ET-1, a gene product involved in heart failure and whose inhibition improves the survival rate of rats with heart failure. In a failing heart, a metabolic switch occurs, and HIF1α activates the expression of glycolytic enzymes as compensation for impaired b-oxidation of fatty acid. Another consequence of increased HIF1α activity is that in rat cardiomyocytes, HIF1α was shown to bind to the 5′-promoter region of the ET-1 gene and increase ET-1 expression. In vitro, an antisense oligonucleotide targeted to hypoxia-inducible factor-1 alphalargely inhibited the increased gene expression of ET-1, confirming the role of HIF1α in heart failure (Kakinuma et al., Circulation, 2001, 103, 2387-2394). This antisense oligonucleotide is comprised of 20 nucleotides and targets bases 11 to 31 of the rat HIF1α with GenBank accession number AF—057308 incorporated herein by reference.
Preeclampsia is a disorder of unknown etiology that is the leading cause of fetal and maternal morbidity and mortality. Defective downregulation of HIF1α may play a major role in the pathogenesis of preeclampsia. For most of the first trimester, the human fetus develops under hypoxic conditions but at 10-12 weeks the intervillous space opens, the fetus is exposed to maternal blood and at this stage the trophoblast cells invade the maternal decidua. The switch of the trophoblasts from a proliferative to an invasive phenotype is controlled by cellular oxygen concentration. The proliferative, non-invasive trophoblast phenotype appears to be maintained by HIF1α mediated expression of TGFbeta3 because treatment of human villous explants with an antisense oligonucleotide against HIF1α or TGF beta 3 induces invasion under hypoxic conditions. In this case the HIF1α antisense oligonucleotide was comprised of phosphorothioate oligonucleotides, 16 nucleotides in length, and targeted to the AUG codon (Caniggia et al., J. Clin. Invest., 2000, 105, 577-587;. Caniggia et al., Placenta, 2000, 21 Suppl A, S25-30).
The human intestinal trefoil factor (ITF) gene product protects the epithelial barrier during episodes of intestinal hypoxia. The ITF gene promoter bears a binding site for hypoxia-inducible factor-1 alpha, and the function of HIF1α as a transcription factor for ITF was confirmed in vitro. T84 colonic epithelial cells were treated with a phosphorothioate antisense oligonucleotide, 15 nucleotides in length and targeted to the AUG codon of HIF1α and this resulted in a loss of ITF hypoxia inducibility (Furuta et al., J. Exp. Med., 2001, 193, 1027-1034).
Human epidemiological and animal studies have associated inhalation of nickel dusts with an increased incidence of pulmonary fibrosis. Nickel transcriptionally activates plasminogen activator inhibitor (PAI-1), an inhibitor of fibrinolysis, through the HIF1α signaling pathway. This was evidenced by decreases in PAI-1 mRNA levels when human airway epithelial cells were treated with an antisense oligonucleotide directed against HIF1α identical to the one used in the preeclampsia study discussed above. These data may be critical for understanding the pathology of pulmonary fibrosis and other diseases associated with nickel exposure (Andrew et al., Am J Physiol Lung Cell Mol Physiol, 2001, 281, L607-615).
HIF1α is constitutively expressed in cerebral neurons under normoxic conditions. A second dimerization partner for HIF1α is ARNT2, a cerebral translocator homologous to hypoxia-inducible factor-1 beta. One splice variant of HIF1α found in rat neurons dimerizes with ARNT2 more avidly than it does with HIF1b, and the resulting hypoxia-inducible factor-1 alpha-ARNT2 heterodimer does not recognize the HIF1α binding site of the erythropoietin gene. This suggests that transcription of a different set of genes is controlled by the hypoxia-inducible factor-1 alpha-ARNT2 heterodimer controls in neurons under nonhypoxic conditions than the hypoxia-inducible factor-1 alpha-HIF1α heterodimer controls under hypoxic conditions. This was evidenced by antisense oligonucleotide downregulation of HIF1α expression in which the antisense oligonucleotide consisted of 16 phosphorothioate nucleotides targeted to bases 38 to 54 of the rat hypoxia-inducible factor-1 with GenBank accession number AF—057308 (Drutel et al., Eur. J. Neurosci., 2000, 12, 3701-3708).
A role for HIF1α in mediating a down-regulatory pathway was recently discovered using antisense oligonucleotide depletion of hypoxia-inducible factor-1 alpha. The peroxisome proliferator-activated receptors (PPARs) are a nuclear hormone-binding proteins that regulate transcriptional activities. Ligands which bind the PPAR-gamma isoform man amplify or inhibit the expression of inflammation-related gene products and may regulate the duration of inflammatory response. Hypoxia elicits a down-regulation of PPAR-gamma in intestinal epithelial cells which is effected through a binding site for HIF1α on the antisense strand of the PPAR-gamma gene. The expression of PPAR-gamma was upregulated in hypoxic cells when treated with an antisense oligonucleotide targeted to HIF1α identical to the one used in the preeclampsia study discussed above (Narravula and Colgan, J. Immunol., 2001, 166, 7543-7548).
The gene encoding hypoxia-inducible factor 2 alpha (HIF2α; also called HIF-2 alpha, endothelial PAS domain protein 1, EPAS1, MOP2, hypoxia-inducible factor 2, HIF-related factor, HRF, HIF1 alpha-like factor, HLF) was initially identified as a transcription factor expressed in endothelial cells (Ema et al., Proc. Natl. Acad. Sci. U.S.A., 1997, 94, 4273-4278; Flamme et al., Mech. Dev., 1997, 63, 51-60; Hogenesch et al., J. Biol. Chem., 1997, 272, 8581-8593; Tian et al., Genes Dev., 1997, 11, 72-82). A nucleic acid sequence encoding human HIF2α is disclosed and claimed in U.S. Pat. No. 5,695,963 (McKnight et al., 1997).
HIF2α mRNA is primarily expressed in highly vascularized adult tissues, such as lung, heart and liver, and in the placenta and endothelial cells of the embryonic and adult mouse (Hogenesch et al., J. Biol. Chem., 1997, 272, 8581-8593). Comparison of normal human tissues and cancers reveals that HIF2α protein is not detectable in normal tissue, but is easily visualized in malignant tissues (Talks et al., Am. J. Pathol., 2000, 157, 411-421). The requirement for expression of HIF2α in development is demonstrated by the abnormalities observed in HIF2α gene deficient mouse embryos, which include the disruption of catecholamine homeostasis and lack of protection against heart failure observed (Tian et al., Genes Dev., 1998, 12, 3320-3324). Targeted disruption of the HIF2α gene and generation of embryos deficient for HIF2α is disclosed in the PCT publication WO 02/086497 (Compernolle et al., 2002). This publication also discloses antisense oligodeoxyribonucleotides for use in inhibiting HIF2α expression targeted to the translation initiation codon of HIF2α (Compernolle et al., 2002).
HIF2α expression and HIF transcriptional activity are precisely regulated by cellular oxygen concentration. Whereas changes in oxygen levels do not affect HIF1-beta protein levels, the abundance of the alpha subunits is markedly increased upon exposure of cells to hypoxia, primarily due to stabilization of the alpha subunit protein (Safran and Kaelin, J. Clin. Invest., 2003, 111, 779-783). HIF2α mRNA and protein is expressed at low levels in tissue culture cells, but protein expression is markedly induced by exposure to 1% oxygen, a hypoxic state (Wiesener et al., Blood, 1998, 92, 2260-2268). The hypoxia-inducible factor 2 alpha/hypoxia-inducible factor 1 beta heterodimer protein binds to the hypoxic response element, which contains the core recognition sequence 5′-TACGTG-3′ and is found in the cis-regulatory regions of hypoxia-regulated genes (Ema et al., Proc. Natl. Acad. Sci. U.S.A., 1997, 94, 4273-4278; Hogenesch et al., J. Biol. Chem., 1997, 272, 8581-8593). Binding of the heterodimer to the HRE induces gene expression. Upon return to normoxic conditions, HIF2α protein is rapidly degraded (Wiesener et al., Blood, 1998, 92, 2260-2268).
The mitogen-activated protein kinase (MAPK) pathway is critical for HIF2α activation. Inhibition of a dual specificity protein kinase that directly phosphorylates MAPK prevents HIF2α trans-activation during hypoxia (Conrad 1999; Conrad, 2001). However, the inhibitor does not prevent HIF2α phosphorylation, thus, while the MAPK pathway regulates the activity of hypoxia-inducible factor 2 alpha, it does not directly phosphorylate the protein (Conrad et al., Comp. Biochem. Physiol. B. Biochem. Mol Biol., 2001, 128, 187-204; Conrad et al., J. Biol. Chem., 1999, 274, 33709-33713). The Src family kinase pathway is also implicated in regulation of hypoxia-inducible factor 2 alpha. A specific inhibitor of the Src family of kinases abolishes the hypoxia-induced expression of HIF2α mRNA in human lung adenocarcinoma cells (Sato et al., Am. J. Respir. Cell Mol. Biol., 2002, 26, 127-134).
The maintenance of oxygen homeostasis, in addition to being required in physiological development, is also required in tumor growth. Tumor cells experience hypoxia because blood circulates poorly through the aberrant blood vessel that tumors establish. Although hypoxia is toxic to cancer cells, they survive as a result of genetic and adaptive changes that allow them to thrive in a hypoxic environment. One such adaptation is an increase in the expression of the angiogenic growth factor named vascular endothelial growth factor (VEGF). VEGF is a key angiogenic factor secreted by cancer cells, as well as normal cells, in response to hypoxia (Harris, Nat. Rev. Cancer, 2002, 2, 38-47; Maxwell et al., Curr. Opin. Genet. Dev., 2001, 11, 293-299).
Hemangioblastomas, the most frequent manifestation of VHL gene mutations, exhibit overexpression of VEGF mRNA in their associated stromal cells. The VEGF mRNA overexpression is highly correlated with elevated expression of HIF2α mRNA. This finding suggests a relationship between loss of function of the VHL gene, and transcriptional activation of the VEGF gene, possibly through HIF2α activity in VEGF-dependent vascular growth (Flamme et al., Am. J. Pathol., 1998, 153, 25-29).
The tumor suppressive activity of the VHL gene product can be overridden by the activation of HIF target genes in human renal carcinoma cells in vivo. VHL gene product mutants lose the ability to target HIF for ubiquitin-mediated destruction, suggesting that down regulation of HIF and VHL tumor suppressor function are intimately linked (Kondo et al., Cancer Cell, 2002, 1, 237-246). In contrast to human renal cell carcinoma, the product of the tuberous sclerosis complex-2 (Tsc-2) gene, product rather than VEIL gene, is the primary target for rodent renal cell carcinoma (Liu et al., Cancer Res., 2003, 63, 2675-2680). Rat RCC cells lacking Tsc-2 function exhibit stabilization of HIF2α protein and upregulation of VEGF, and were highly vascularized (Liu et al., Cancer Res., 2003, 63, 2675-2680).
A link between elevated HIF2α activity and angiogenesis has also been demonstrated by experiments that show how HIF activity regulates VEGF expression. Normal human kidney cells typically have low levels of hypoxia-inducible factor 2 alpha, but upon introduction of a vector encoding HIF2α into these cells, VEGF mRNA and protein levels increase significantly (Xia et al., Cancer, 2001, 91, 1429-1436). When HIF2α was inhibited, VEGF expression was significantly decreased, thus demonstrating a direct link between HIF2α activity and VEGF expression (Xia et al., Cancer, 2001, 91, 1429-1436). Similarly, a dose-dependent increase in VEGF mRNA is observed when human umbulical vein cells are transduced with a virus encoding HIF2α (Maemura et al., J. Biol. Chem., 1999, 274, 31565-31570). Expression of a mutated HIF2α that lacks a transactivation domain inhibits the induction of VEGF mRNA during hypoxia, a finding that further suggests that HIF2α is an important regulator of VEGF expression (Maemura et al., J. Biol. Chem., 1999, 274, 31565-31570).
A correlation between HIF activity and VEGF expression is also observed in malignant cells and tissues. HIF2α can be readily detected in renal cell carcinoma (RCC) cell lines in the absence of a vector encoding HIF2α (Xia et al., Cancer, 2001, 91, 1429-1436). Significant increases in HIF2α and VEGF mRNA in renal cell carcinoma tissue samples, compared to normal tissue, suggest that abnormal activation of HIF2α may be involved in the angiogenesis of RCC (Xia et al., Cancer, 2001, 91, 1429-1436).
In addition to RCC, the expression of HIF2α in other malignancies has also been reported. HIF2α is expressed at the levels of mRNA and protein in human bladder cancers, especially in those with an invasive phenotype (Xia et al., Urology, 2002, 59, 774-778). Another example of overexpression of HIF2α is seen in squamous cell head-and-neck cancer (SCHNC). Higher levels of HIF2α were associated with locally aggressive behavior of SCHNC, as well as intensification of angiogenesis (Koukourakis et al., Int. J. Radiat. Oncol. Biol. Phys., 2002, 53, 1192-1202). These findings also demonstrated a link between overexpression of HIF2α and resistance to chemotherapy. Yet another correlation between overexpression of HIF2α and cancer is seen in malignant pheochromocytomas, which exhibit a higher level of HIF2α and an induced VEGF pathway, when compared to benign counterparts (Favier et al., Am. J. Pathol., 2002, 161, 1235-1246). HIF2α overexpression is also a common event in non-small-cell lung cancer (NSCLC) and is related to the up-regulation of multiple angiogenic factors and overexpression of angiogenic receptors by cancer cells. HIF2α overexpression in NSCLC is an indicator of poor prognosis (Giatromanolaki et al., Br. J. Cancer, 2001, 85, 881-890). Taken together, these studies demonstrate that elevated HIF2α confers aggressive tumor behavior, and that targeting the HIF pathway may aid the treatment of several different types of cancers.
Overexpression of HIF2α has also been observed in several cancer cell lines in addition to RCC cell lines. Elevated levels of HIF2α mRNA and protein are seen in human lung adenocarcinoma cells, and exposure of these cells to hypoxia further increases HIF2α expression (Sato et al., Am. J. Respir. Cell Mol. Biol., 2002, 26, 127-134). Furthermore, the hypoxia response element plays a role in constitutively upregulating an isoform of VEGF in cancer cell lines under normoxic conditions. The HRE located within a cell type-specific enhancer element in glioblastoma cells participates in the up-regulation of VEGF expression through enhanced binding of HIF2α to the HRE (Liang et al., J. Biol. Chem., 2002, 277, 20087-20094). A truncated version of HIF2α that can bind to hypoxia-inducible factor 1 beta, but not to the HRE, was unable to transactivate the VEGF promoter (Liang et al., J. Biol. Chem., 2002, 277, 20087-20094). This further demonstrates the capability of cancer cells to combat hypoxic conditions by enhancing expression of factors required for vascularization and angiogenesis.
Short interfering RNAs (siRNAs) have been used to specifically inhibit the expression of HIF1α and HIF2α in human breast and renal carcinoma cell lines and in a human endothelial cell line. SiRNA duplexes with dTdT overhangs at both ends were designed to target nucleotides 1521-1541 and 1510-1530 of the HIF1α mRNA sequence (NM001530) and nucleotides 1260-1280 and 328-348 of the HIF2α sequence (NM001430). It was found that in the breast carcinoma and endothelial cell lines, gene expression and cell migration patterns were critically dependent on HIF1α but not hypoxia-inducible factor-2 alpha, but critically dependent on HIF2α in the renal carcinoma cells. Sowter et al., 2003, Cancer Res., 63, 6130-6134.
Defective downregulation of HIF2α may play a major role in the pathogenesis of preeclampsia. HIF2α protein levels are increased during early development, as expected in a hypoxic environment, and then decrease significantly with gestational age (Rajakumar and Conrad, Biol. Reprod., 2000, 63, 559-569). However, HIF2α protein expression is significantly increased in preeclamptic relative to normal term placentas (Rajakumar et al., Biol. Reprod., 2001, 64, 499-506). This result suggests that failure to down-regulate HIF2α protein expression during early pregnancy could prevent the switch of the trophoblast to an invasive phenotype and ultimately lead to preeclampsia (Rajakumar et al., Biol. Reprod, 2001, 64, 499-506).
Overexpression of hypoxia-inducible factor 2 alpha, as well as hypoxia-inducible factor 1, has been observed in the inflammatory bowel diseases Crohn's disease and ulcerative colitis (Giatromanolaki et al., J. Clin. Pathol., 2003, 56, 209-213). However, VEGF expression was weak in ulcerative colitis samples, and absent in Crohn's disease samples. The discordant expression of VEGF and HIF2α may lead to a reduced ability of a tissue to produce or respond to VEGF, which may in turn lead to reduced endothelial and epithelial cell viability (Giatromanolaki et al., J. Clin. Pathol., 2003, 56, 209-213).
In addition to participating in adaptive changes in response to hypoxia, HIF2α may also function in an inflammatory response in cardiac myocytes. In cultured cardiac myocytes, interleukin-1 beta (IL-1beta) significantly increased both HIF2α mRNA and protein levels (Tanaka et al., J. Mol. Cell Cardiol., 2002, 34, 739-748). Transduction of cardiac myocytes with adenovirus expressing HIF2α dramatically increased the levels of adrenomedullin (AM) mRNA, which is also upregulated by IL-1beta (Tanaka et al., J. Mol. Cell Cardiol., 2002, 34, 739-748). Since IL-1 beta has been implicated in the pathogenesis of heart failure, and AM is known to improve cardiac function during heart failure, these results suggest that HIF2α plays a role in the adaptation of the cardiac myocytes during heart failure (Tanaka et al., J. Mol. Cell Cardiol., 2002, 34, 739-748).
Disclosed and claimed in the PCT publication WO 00/09657 is a method of inhibiting angiogenesis in a mammal through administration of a compound which inhibits the binding of human HIF2α protein to the DNA regulatory element of an angiogenic factor, wherein the compound can be an antisense nucleic acid molecule complementary to all or part of the mRNA encoding human HIF2α (Lee et al., 2000). This publication also discloses a nucleic acid encoding human hypoxia-inducible factor 2 alpha.
The PCT publication WO 01/62965 discloses and claims a differential screening method for identifying a genetic element involved in a cellular process, which method includes introducing HIF2α into cells (Kingsman, 2001). This publication also discloses the development of HIF2α agonists or antagonists.
The PCT publication WO 02/34291 claims methods and reagents, including the use of antisense oligonucleotides, for the inhibition of human HIF1α transcription (Colgan, 2002). This publication also discloses a nucleic acid encoding human hypoxia-inducible factor 2 alpha.
U.S. Pat. No. 6,395,548 claims a nucleic acid encoding a deletion mutant of human HIF2α and the use of this deletion mutant as a method of inhibiting expression of an angiogenic factor in vitro. This patent also discloses a nucleic acid encoding human hypoxia-inducible factor 2 alpha, as well as nucleic acids complementary to all or part of the human HIF2α cDNA for use in antisense treatment to inhibit the expression of HIF2α (Lee et al., 2002).
U.S. Pat. No. 6,432,927 discloses nucleic acid sequences, including sense and antisense oligonucleotides, which are derived from an HIF2α and incorporated into recombinant nucleic acid molecules for the purpose of sustaining HIF2α expression in cells (Gregory and Vincent, 2002).
The nucleic acid sequence encoding a human HIF2α and insertion of this sequence into a viral expression vector, for the purpose of driving human HIF2α expression in mammalian cells, is disclosed in the PCT publication WO 02/068466 (White et al., 2002).
The PCT publication WO 02/094862 discloses a method for introducing into a muscle cell a nucleic acid sequence encoding hypoxia-inducible factor 2 alpha, for the purpose of overexpressing HIF2α and stimulating angiogenesis or metabolic activity (Guy, 2002).
Disclosed and claimed in the US pre-grant publication 2003/0045686 is a nucleic acid encoding human hypoxia-inducible factor 2 alpha, and a method of delivering a therapeutically effective amount of this nucleic acid to a subject for the purpose of reducing or preventing hypoxia (Kaelin Jr. and Ivan, 2003). This publication also discloses and claims human HIF muteins, including HIF2α mutein, which are designed to be more stable and/or resistant to degradation.
As a consequence of HIF2α involvement in many diseases, there remains a long felt need for additional agents capable of effectively regulating HIF2α function. As such, inhibition is especially important in the treatment of cancer, given that the upregulation of expression of HIF2α is associated with so many different types of cancer.
As a consequence of HIF1α and HIF2α involvement in many diseases, there remains a long felt need for additional agents capable of effectively inhibiting HIF1α and HIF2α function.
Antisense technology is emerging as an effective means for reducing the expression of specific gene products and may therefore prove to be uniquely useful in a number of therapeutic, diagnostic, and research applications for the modulation of HIF1α and HIF2α expression.
The present invention provides compositions and methods for modulating HIF1α and HIF2α expression. In particular antisense compositions for modulating HIF1α and/or HIF2α expression are believed to be useful in treatment of abnormal proliferative conditions associated with HIF1α and/or HIF2α. Examples of abnormal proliferative conditions are hyperproliferative disorders such as cancers, tumors, hyperplasias, pulmonary fibrosis, angiogenesis, psoriasis, atherosclerosis and smooth muscle cell proliferation in the blood vessels, such as stenosis or restenosis following antioplasty. It is presently believed that inhibition of both HIF1α and HIF2α may be a particularly useful approach to treatment of such disorders.